[unreadable] Rheumatoid arthritis (RA) is a systemic inflammatory disease that leads to progressive joint destruction, [unreadable] disability, and accelerated mortality. Factors that influence autoantibody production and disease everity in RA have not been fully defined. However, there is increasing data showing that cigarette smoking, [unreadable] associated with disease susceptibility, is also associated with RA outcome and its effect is likely modified by multiple genetic factors including HLA-DRB1 alleles encoding the shared epitope (SE) and drug metabolizing enzyme (DME) polymorphisms. Studies examining RA severity have focused on the [unreadable] interaction of smoking with single gene polymorphisms in groups that have almost exclusively included [unreadable] Caucasian women. This is an important distinction because smoking, in terms of RA risk (particularly for autoantibody positive disease), has its greatest impact in men while other smoking related illnesses [unreadable] disproportionately impact non-Caucasians. We will examine determinants of autoantibody production and disease severity in 800 subjects (including 600 Caucasian men) from the Veterans Affairs RA (VARA) cohort and 400 African Americans from the NIH-funded Consortium for the Longitudinal Evaluation of African Americans with Early RA (CLEAR) registry. The overall hypothesis of this study is that smoking is associated with greater autoantibody production and more severe RA and the effect of this environmental exposure is modified by multiple gene-smoking interactions. The aims of this study are to examine: 1) associations of smoking with autoantibody levels and radiographic measures, 2) associations of DME polymorphisms with these outcomes, and 3) the role of gene-smoking interactions relevant to these outcomes. In addition to SE, genetic factors to be studied will include DME polymorphisms (in genes encoding N-acetyltransferase [NAT]1, NAT2, microsomal epoxide hydrolase, and glutathione S-transferase), and protein tyrosine phosphatase (PTPN22). RA-specific outcomes that will be examined include radiographic measures (modified Sharp score), nodules, rheumatoid factor (RF), and anti-cyclic citrullinated peptide (CCP) antibody. In addition to traditional statistical approaches, a novel recursive partitioning technique will be used that will allow for the detection simultaneous interactions among multiple candidate genes and smoking. We anticipate that the results of this study will be extended to other RA populations. [unreadable] [unreadable] [unreadable]